Merge bk://kernel.bkbits.net/davem/net-2.6

into home.osdl.org:/home/torvalds/v2.5/linux

15 files changed, 141 insertions, 75 deletions

diff --git a/net/appletalk/sysctl_net_atalk.c b/net/appletalk/sysctl_net_atalk.c
index edddd3291f70..25b33f670499 100644
--- a/net/appletalk/sysctl_net_atalk.c
+++ b/net/appletalk/sysctl_net_atalk.c
@@ -23,6 +23,7 @@ static struct ctl_table atalk_table[] = {
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies,
 	},
 	{
 		.ctl_name	= NET_ATALK_AARP_TICK_TIME,
@@ -31,6 +32,7 @@ static struct ctl_table atalk_table[] = {
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies,
 	},
 	{
 		.ctl_name	= NET_ATALK_AARP_RETRANSMIT_LIMIT,
@@ -47,6 +49,7 @@ static struct ctl_table atalk_table[] = {
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies,
 	},
 	{ 0 },
 };
diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c
index 932f3ae2bbea..359555cd35b6 100644
--- a/net/bridge/br_netfilter.c
+++ b/net/bridge/br_netfilter.c
@@ -356,6 +356,7 @@ static unsigned int br_nf_local_in(unsigned int hook, struct sk_buff **pskb,
 	return NF_ACCEPT;
 }
 
+
 /* PF_BRIDGE/FORWARD *************************************************/
 static int br_nf_forward_finish(struct sk_buff *skb)
 {
@@ -466,6 +467,7 @@ static unsigned int br_nf_forward_arp(unsigned int hook, struct sk_buff **pskb,
 	return NF_STOLEN;
 }
 
+
 /* PF_BRIDGE/LOCAL_OUT ***********************************************/
 static int br_nf_local_out_finish(struct sk_buff *skb)
 {
@@ -531,9 +533,7 @@ static unsigned int br_nf_local_out(unsigned int hook, struct sk_buff **pskb,
 		return NF_ACCEPT;
 
 	nf_bridge = skb->nf_bridge;
-
 	nf_bridge->physoutdev = skb->dev;
-
 	realindev = nf_bridge->physindev;
 
 	/* Bridged, take PF_BRIDGE/FORWARD.
@@ -601,18 +601,15 @@ static unsigned int br_nf_post_routing(unsigned int hook, struct sk_buff **pskb,
 	struct vlan_ethhdr *hdr = (struct vlan_ethhdr *)(skb->mac.ethernet);
 	struct net_device *realoutdev = bridge_parent(skb->dev);
 
-	/* Be very paranoid. Must be a device driver bug. */
+#ifdef CONFIG_NETFILTER_DEBUG
+	/* Be very paranoid. This probably won't happen anymore, but let's
+	 * keep the check just to be sure... */
 	if (skb->mac.raw < skb->head || skb->mac.raw + ETH_HLEN > skb->data) {
 		printk(KERN_CRIT "br_netfilter: Argh!! br_nf_post_routing: "
 				 "bad mac.raw pointer.");
-		if (skb->dev != NULL) {
-			printk("[%s]", skb->dev->name);
-			if (has_bridge_parent(skb->dev))
-				printk("[%s]", bridge_parent(skb->dev)->name);
-		}
-		printk(" head:%p, raw:%p\n", skb->head, skb->mac.raw);
-		return NF_ACCEPT;
+		goto print_error;
 	}
+#endif
 
 #ifdef CONFIG_SYSCTL
 	if (!nf_bridge)
@@ -622,13 +619,16 @@ static unsigned int br_nf_post_routing(unsigned int hook, struct sk_buff **pskb,
 	if (skb->protocol != __constant_htons(ETH_P_IP) && !IS_VLAN_IP)
 		return NF_ACCEPT;
 
+#ifdef CONFIG_NETFILTER_DEBUG
 	/* Sometimes we get packets with NULL ->dst here (for example,
-	 * running a dhcp client daemon triggers this).
+	 * running a dhcp client daemon triggers this). This should now
+	 * be fixed, but let's keep the check around.
 	 */
-	if (skb->dst == NULL)
-		return NF_ACCEPT;
+	if (skb->dst == NULL) {
+		printk(KERN_CRIT "br_netfilter: skb->dst == NULL.");
+		goto print_error;
+	}
 
-#ifdef CONFIG_NETFILTER_DEBUG
 	skb->nf_debug ^= (1 << NF_IP_POST_ROUTING);
 #endif
 
@@ -655,6 +655,18 @@ static unsigned int br_nf_post_routing(unsigned int hook, struct sk_buff **pskb,
 		realoutdev, br_dev_queue_push_xmit);
 
 	return NF_STOLEN;
+
+#ifdef CONFIG_NETFILTER_DEBUG
+print_error:
+	if (skb->dev != NULL) {
+		printk("[%s]", skb->dev->name);
+		if (has_bridge_parent(skb->dev))
+			printk("[%s]", bridge_parent(skb->dev)->name);
+	}
+	printk(" head:%p, raw:%p, data:%p\n", skb->head, skb->mac.raw,
+					      skb->data);
+	return NF_ACCEPT;
+#endif
 }
 
 
diff --git a/net/core/dev.c b/net/core/dev.c
index 47fff21ea3b2..30cba1e1c633 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1543,7 +1543,7 @@ static inline int __handle_bridge(struct sk_buff *skb,
 			struct packet_type **pt_prev, int *ret)
 {
 #if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
-	if (skb->dev->br_port) {
+	if (skb->dev->br_port && skb->pkt_type != PACKET_LOOPBACK) {
 		*ret = handle_bridge(skb, *pt_prev);
 		if (br_handle_frame_hook(skb) == 0)
 			return 1;
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index 247b6d6e661d..4b194b6be05a 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -1518,6 +1518,7 @@ struct neigh_sysctl_table {
 			.maxlen		= sizeof(int),
 			.mode		= 0644,
 			.proc_handler	= &proc_dointvec_jiffies,
+			.strategy	= &sysctl_jiffies,
 		},
 		{
 			.ctl_name	= NET_NEIGH_DELAY_PROBE_TIME,
@@ -1525,6 +1526,7 @@ struct neigh_sysctl_table {
 			.maxlen		= sizeof(int),
 			.mode		= 0644,
 			.proc_handler	= &proc_dointvec_jiffies,
+			.strategy	= &sysctl_jiffies,
 		},
 		{
 			.ctl_name	= NET_NEIGH_GC_STALE_TIME,
@@ -1532,6 +1534,7 @@ struct neigh_sysctl_table {
 			.maxlen		= sizeof(int),
 			.mode		= 0644,
 			.proc_handler	= &proc_dointvec_jiffies,
+			.strategy	= &sysctl_jiffies,
 		},
 		{
 			.ctl_name	= NET_NEIGH_UNRES_QLEN,
@@ -1574,6 +1577,7 @@ struct neigh_sysctl_table {
 			.maxlen		= sizeof(int),
 			.mode		= 0644,
 			.proc_handler	= &proc_dointvec_jiffies,
+			.strategy	= &sysctl_jiffies,
 		},
 		{
 			.ctl_name	= NET_NEIGH_GC_THRESH1,
diff --git a/net/core/sock.c b/net/core/sock.c
index 9d51516eaa56..d549d519482d 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -1119,6 +1119,31 @@ void sock_init_data(struct socket *sock, struct sock *sk)
 	atomic_set(&sk->sk_refcnt, 1);
 }
 
+void lock_sock(struct sock *sk)
+{
+	might_sleep();
+	spin_lock_bh(&(sk->sk_lock.slock));
+	if (sk->sk_lock.owner)
+		__lock_sock(sk);
+	sk->sk_lock.owner = (void *)1;
+	spin_unlock_bh(&(sk->sk_lock.slock));
+}
+
+EXPORT_SYMBOL(lock_sock);
+
+void release_sock(struct sock *sk)
+{
+	spin_lock_bh(&(sk->sk_lock.slock));
+	if (sk->sk_backlog.tail)
+		__release_sock(sk);
+	sk->sk_lock.owner = NULL;
+        if (waitqueue_active(&(sk->sk_lock.wq)))
+		wake_up(&(sk->sk_lock.wq));
+	spin_unlock_bh(&(sk->sk_lock.slock));
+}
+
+EXPORT_SYMBOL(release_sock);
+
 EXPORT_SYMBOL(__lock_sock);
 EXPORT_SYMBOL(__release_sock);
 EXPORT_SYMBOL(sk_alloc);
diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c
index 46c0c78b196d..642ef77a2824 100644
--- a/net/core/sysctl_net_core.c
+++ b/net/core/sysctl_net_core.c
@@ -146,7 +146,8 @@ ctl_table core_table[] = {
 		.data		= &net_msg_cost,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec_jiffies
+		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies,
 	},
 	{
 		.ctl_name	= NET_CORE_MSG_BURST,
@@ -154,7 +155,8 @@ ctl_table core_table[] = {
 		.data		= &net_msg_burst,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
-		.proc_handler	= &proc_dointvec_jiffies
+		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies,
 	},
 	{
 		.ctl_name	= NET_CORE_OPTMEM_MAX,
diff --git a/net/ipv4/netfilter/ipt_recent.c b/net/ipv4/netfilter/ipt_recent.c
index 3e03fe295529..eda86a22accd 100644
--- a/net/ipv4/netfilter/ipt_recent.c
+++ b/net/ipv4/netfilter/ipt_recent.c
@@ -91,8 +91,10 @@ static struct recent_ip_tables *r_tables = NULL;
  */
 static spinlock_t recent_lock = SPIN_LOCK_UNLOCKED;
 
+#ifdef CONFIG_PROC_FS
 /* Our /proc/net/ipt_recent entry */
 static struct proc_dir_entry *proc_net_ipt_recent = NULL;
+#endif
 
 /* Function declaration for later. */
 static int
@@ -959,8 +961,10 @@ static int __init init(void)
 	int count;
 
 	printk(version);
+#ifdef CONFIG_PROC_FS
 	proc_net_ipt_recent = proc_mkdir("ipt_recent",proc_net);
 	if(!proc_net_ipt_recent) return -ENOMEM;
+#endif
 
 	if(ip_list_hash_size && ip_list_hash_size <= ip_list_tot) {
 	  printk(KERN_WARNING RECENT_NAME ": ip_list_hash_size too small, resetting to default.\n");
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index b08ba41c5356..0d4b1688a801 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -2946,6 +2946,7 @@ static struct addrconf_sysctl_table
 			.maxlen		=	sizeof(int),
 			.mode		=	0644,
          		.proc_handler	=	&proc_dointvec_jiffies,
+			.strategy	=	&sysctl_jiffies,
 		},
 		{
 			.ctl_name	=	NET_IPV6_RTR_SOLICIT_DELAY,
@@ -2954,6 +2955,7 @@ static struct addrconf_sysctl_table
 			.maxlen		=	sizeof(int),
 			.mode		=	0644,
          		.proc_handler	=	&proc_dointvec_jiffies,
+			.strategy	=	&sysctl_jiffies,
 		},
 #ifdef CONFIG_IPV6_PRIVACY
 		{
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 0309a24a367e..f2e7c8bb7fe9 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -75,6 +75,9 @@
 #include <net/checksum.h>
 #include <linux/proc_fs.h>
 
+#include <linux/netfilter.h>
+#include <linux/netfilter_ipv6.h>
+
 static struct socket *ndisc_socket;
 
 static u32 ndisc_hash(const void *pkey, const struct net_device *dev);
@@ -497,10 +500,11 @@ static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
 
 	skb->dst = dst;
 	idev = in6_dev_get(dst->dev);
-	dst_output(skb);
-
-	ICMP6_INC_STATS(idev, Icmp6OutNeighborAdvertisements);
-	ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	err = NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev, dst_output);
+	if (!err) {
+		ICMP6_INC_STATS(idev, Icmp6OutNeighborAdvertisements);
+		ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	}
 
 	if (likely(idev != NULL))
 		in6_dev_put(idev);
@@ -576,10 +580,11 @@ void ndisc_send_ns(struct net_device *dev, struct neighbour *neigh,
 	/* send it! */
 	skb->dst = dst;
 	idev = in6_dev_get(dst->dev);
-	dst_output(skb);
-
-	ICMP6_INC_STATS(idev, Icmp6OutNeighborSolicits);
-	ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	err = NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev, dst_output);
+	if (!err) {
+		ICMP6_INC_STATS(idev, Icmp6OutNeighborSolicits);
+		ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	}
 
 	if (likely(idev != NULL))
 		in6_dev_put(idev);
@@ -644,10 +649,11 @@ void ndisc_send_rs(struct net_device *dev, struct in6_addr *saddr,
 	/* send it! */
 	skb->dst = dst;
 	idev = in6_dev_get(dst->dev);
-	dst_output(skb);
-
-	ICMP6_INC_STATS(idev, Icmp6OutRouterSolicits);
-	ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	err = NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev, dst_output);
+	if (!err) {
+		ICMP6_INC_STATS(idev, Icmp6OutRouterSolicits);
+		ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	}
 
 	if (likely(idev != NULL))
 		in6_dev_put(idev);
@@ -1404,10 +1410,11 @@ void ndisc_send_redirect(struct sk_buff *skb, struct neighbour *neigh,
 
 	buff->dst = dst;
 	idev = in6_dev_get(dst->dev);
-	dst_output(buff);
-
-	ICMP6_INC_STATS(idev, Icmp6OutRedirects);
-	ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	err = NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, buff, NULL, dst->dev, dst_output);
+	if (!err) {
+		ICMP6_INC_STATS(idev, Icmp6OutRedirects);
+		ICMP6_INC_STATS(idev, Icmp6OutMsgs);
+	}
 
 	if (likely(idev != NULL))
 		in6_dev_put(idev);
diff --git a/net/sctp/associola.c b/net/sctp/associola.c
index dba3d149e499..44b30efb17a9 100644
--- a/net/sctp/associola.c
+++ b/net/sctp/associola.c
@@ -58,6 +58,7 @@
 #include <linux/in.h>
 #include <net/ipv6.h>
 #include <net/sctp/sctp.h>
+#include <net/sctp/sm.h>
 
 /* Forward declarations for internal functions. */
 static void sctp_assoc_bh_rcv(struct sctp_association *asoc);
diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c
index d26292af9ff1..dc4cb1a5a739 100644
--- a/net/sctp/outqueue.c
+++ b/net/sctp/outqueue.c
@@ -53,6 +53,7 @@
 #include <net/sock.h>	  /* For skb_set_owner_w */
 
 #include <net/sctp/sctp.h>
+#include <net/sctp/sm.h>
 
 /* Declare internal functions here.  */
 static int sctp_acked(struct sctp_sackhdr *sack, __u32 tsn);
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index 0ac8c9bf5363..36b2ec11a3a8 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -77,6 +77,7 @@
 #include <linux/socket.h> /* for sa_family_t */
 #include <net/sock.h>
 #include <net/sctp/sctp.h>
+#include <net/sctp/sm.h>
 
 /* WARNING:  Please do not remove the SCTP_STATIC attribute to
  * any of the functions below as they are used to export functions
diff --git a/net/sctp/transport.c b/net/sctp/transport.c
index 485d77015399..026ffebb61d4 100644
--- a/net/sctp/transport.c
+++ b/net/sctp/transport.c
@@ -50,6 +50,7 @@
 
 #include <linux/types.h>
 #include <net/sctp/sctp.h>
+#include <net/sctp/sm.h>
 
 /* 1st Level Abstractions.  */
 
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 85f626d20c4a..0f970906c9ef 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -19,6 +19,8 @@
 #include <linux/list.h>
 #include <linux/spinlock.h>
 #include <linux/workqueue.h>
+#include <linux/notifier.h>
+#include <linux/netdevice.h>
 #include <net/xfrm.h>
 #include <net/ip.h>
 
@@ -690,6 +692,8 @@ static inline int policy_to_flow_dir(int dir)
 	};
 }
 
+static int stale_bundle(struct dst_entry *dst);
+
 /* Main function: finds/creates a bundle for given flow.
  *
  * At the moment we eat a raw IP route. Mostly to speed up lookups
@@ -814,10 +818,11 @@ restart:
 		}
 
 		write_lock_bh(&policy->lock);
-		if (unlikely(policy->dead)) {
+		if (unlikely(policy->dead || stale_bundle(dst))) {
 			/* Wow! While we worked on resolving, this
 			 * policy has gone. Retry. It is not paranoia,
 			 * we just cannot enlist new bundle to dead object.
+			 * We can't enlist stable bundles either.
 			 */
 			write_unlock_bh(&policy->lock);
 
@@ -985,18 +990,27 @@ int __xfrm_route_forward(struct sk_buff *skb, unsigned short family)
 
 static struct dst_entry *xfrm_dst_check(struct dst_entry *dst, u32 cookie)
 {
+	if (!stale_bundle(dst))
+		return dst;
+
+	dst_release(dst);
+	return NULL;
+}
+
+static int stale_bundle(struct dst_entry *dst)
+{
 	struct dst_entry *child = dst;
 
 	while (child) {
 		if (child->obsolete > 0 ||
+		    (child->dev && !netif_running(child->dev)) ||
 		    (child->xfrm && child->xfrm->km.state != XFRM_STATE_VALID)) {
-			dst_release(dst);
-			return NULL;
+			return 1;
 		}
 		child = child->child;
 	}
 
-	return dst;
+	return 0;
 }
 
 static void xfrm_dst_destroy(struct dst_entry *dst)
@@ -1022,7 +1036,7 @@ static struct dst_entry *xfrm_negative_advice(struct dst_entry *dst)
 	return dst;
 }
 
-static void __xfrm_garbage_collect(void)
+static void xfrm_prune_bundles(int (*func)(struct dst_entry *))
 {
 	int i;
 	struct xfrm_policy *pol;
@@ -1034,7 +1048,7 @@ static void __xfrm_garbage_collect(void)
 			write_lock(&pol->lock);
 			dstp = &pol->bundles;
 			while ((dst=*dstp) != NULL) {
-				if (atomic_read(&dst->__refcnt) == 0) {
+				if (func(dst)) {
 					*dstp = dst->next;
 					dst->next = gc_list;
 					gc_list = dst;
@@ -1054,46 +1068,19 @@ static void __xfrm_garbage_collect(void)
 	}
 }
 
-static int bundle_depends_on(struct dst_entry *dst, struct xfrm_state *x)
+static int unused_bundle(struct dst_entry *dst)
 {
-	do {
-		if (dst->xfrm == x)
-			return 1;
-	} while ((dst = dst->child) != NULL);
-	return 0;
+	return !atomic_read(&dst->__refcnt);
 }
 
-int xfrm_flush_bundles(struct xfrm_state *x)
+static void __xfrm_garbage_collect(void)
 {
-	int i;
-	struct xfrm_policy *pol;
-	struct dst_entry *dst, **dstp, *gc_list = NULL;
-
-	read_lock_bh(&xfrm_policy_lock);
-	for (i=0; i<2*XFRM_POLICY_MAX; i++) {
-		for (pol = xfrm_policy_list[i]; pol; pol = pol->next) {
-			write_lock(&pol->lock);
-			dstp = &pol->bundles;
-			while ((dst=*dstp) != NULL) {
-				if (bundle_depends_on(dst, x)) {
-					*dstp = dst->next;
-					dst->next = gc_list;
-					gc_list = dst;
-				} else {
-					dstp = &dst->next;
-				}
-			}
-			write_unlock(&pol->lock);
-		}
-	}
-	read_unlock_bh(&xfrm_policy_lock);
-
-	while (gc_list) {
-		dst = gc_list;
-		gc_list = dst->next;
-		dst_free(dst);
-	}
+	xfrm_prune_bundles(unused_bundle);
+}
 
+int xfrm_flush_bundles(void)
+{
+	xfrm_prune_bundles(stale_bundle);
 	return 0;
 }
 
@@ -1216,6 +1203,21 @@ void xfrm_policy_put_afinfo(struct xfrm_policy_afinfo *afinfo)
 	read_unlock(&afinfo->lock);
 }
 
+static int xfrm_dev_event(struct notifier_block *this, unsigned long event, void *ptr)
+{
+	switch (event) {
+	case NETDEV_DOWN:
+		xfrm_flush_bundles();
+	}
+	return NOTIFY_DONE;
+}
+
+struct notifier_block xfrm_dev_notifier = {
+	xfrm_dev_event,
+	NULL,
+	0
+};
+
 void __init xfrm_policy_init(void)
 {
 	xfrm_dst_cache = kmem_cache_create("xfrm_dst_cache",
@@ -1226,6 +1228,7 @@ void __init xfrm_policy_init(void)
 		panic("XFRM: failed to allocate xfrm_dst_cache\n");
 
 	INIT_WORK(&xfrm_policy_gc_work, xfrm_policy_gc_task, NULL);
+	register_netdevice_notifier(&xfrm_dev_notifier);
 }
 
 void __init xfrm_init(void)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index ab1c1ec1aa12..dc70df79f9cb 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -219,7 +219,7 @@ static void __xfrm_state_delete(struct xfrm_state *x)
 		 * there are DSTs attached to this xfrm_state.
 		 */
 		if (atomic_read(&x->refcnt) > 2)
-			xfrm_flush_bundles(x);
+			xfrm_flush_bundles();
 
 		/* All xfrm_state objects are created by one of two possible
 		 * paths: